HH-polarized Synthetic Aperture Radar Research Articles

Study on Retrievals of Ocean Wave Spectrum by Spaceborne SAR in Ice-Covered Areas

The sea ice in the Arctic is retreating rapidly and ocean waves may accelerate the process by interacting with sea ice. Though Synthetic Aperture Radar (SAR) has shown great capability of imaging waves in ice, there are few attempts to retrieve the ocean wave spectrum (OWS) by SAR in ice-covered areas. In this study, based on the previously developed nonlinear inversion scheme, i.e., the Max Planck Institute (MPI) scheme, and the Sentinel-1 SAR data acquired in the Barents Sea, ocean wave spectra were retrieved by using the different combinations of modulation transfer functions (MTFs) in the MPI scheme: (1) using the same MTFs as those used in open water; (2) by neglecting both the hydrodynamic and tilt modulations; (3) by neglecting the hydrodynamic modulation but involving a newly fitted tilt modulation over ice for HH-polarized SAR data. We compared the simulated SAR image spectra based on the retrievals with the observational SAR image spectra to quantify their respective performances. The comparisons suggest that neglecting hydrodynamic modulation can significantly improve the retrievals. The remaining tilt modulation can further improve the retrievals, particularly for range-travelling waves. This study enhances the understanding of the principles of SAR imaging waves in ice and provides basics for retrievals of ocean wave spectra by SAR data in ice-covered areas.

An automated procedure to map breaking river ice with C-band HH SAR data

The development of effective strategies to manage the river ice breakup process or the associated risks is hindered by a lack of understanding and information. Radar earth observation satellites offer excellent potential for collecting up-to-date information on the conditions of and changes in river ice cover during the breakup period. This text describes the development, performance and limitations of an automated procedure to map breaking river ice by means of C-band, HH-polarized Synthetic Aperture Radar (SAR) images. An original two-step supervised classification model (IceBC), which uses backscatter intensities, lies at the core of the procedure. First, IceBC identifies three primary classes: water, sheet ice and rubble ice. Next, each primary ice class is divided in three secondary classes that denote top surface roughness scale differences. Input images must have incidence angles from ~27° to ~60°. Below ~36°, IceBC may assign a class labelled “unclassified” to water or sheet ice pixels. The primary classification model yields overall accuracies of ~86% and ~93% for independent test pixels with incidence angles ≤ ~49° and ≥ ~29° or ≥ ~36°, respectively. The associated class accuracies for water, sheet ice and rubble ice are ~97% & 96%, ~69% & 85% and ~97% & 99%. Given its connection to ice jam flood events, the classification accuracy achieved for rubble ice is particularly important. Maps produced by means of IceBC comprise detailed spatial information regarding ice cover conditions and the development of the breakup process. Their quality may be affected by: freezing conditions, wet snow cover, meltwater pools, infrastructure, rapids or high winds. Monitoring is the key to managing the impacts of most of these challenges. An IceBC prototype has been used operationally since 2015.

CMODH Validation for C-Band Synthetic Aperture Radar HH Polarization Wind Retrieval Over the Ocean

A new HH-polarized geophysical model function (GMF), called CMODH, has been proposed recently for C-band synthetic aperture radar (SAR) ocean surface wind speed retrieval. CMODH has the potential to be directly used to retrieve wind speed from HH-polarized SAR images without converting the normalized radar cross section (NRCS) from HH- into VV-polarization using various empirical and theoretical polarization ratio (PR) models. However, the capability of CMODH for wind speed retrieval has not been comprehensively validated. In this letter, as a case study, two C-band HH-polarized SAR images acquired in RADARSAT-2 (RS-2) quad-polarization and Sentinel-1A (S1-1A) dual-polarization modes are first used to examine the CMODH performance. The wind speeds estimated using CMODH are shown to be consistent with buoy winds. A statistical comparison is then carried out to further validate CMODH using collocated 1457 C-band RS-2 and 284 S1-1A/B HH-polarized SAR images and 110 in situ buoys. The results show that the CMODH-retrieved wind speeds are in good agreement with buoy measurements, with a bias of 0.07 and 0.49 m/s and a root mean square error (RMSE) of 1.66 and 2.05 m/s for RS-2 and S1-1A/B, respectively. Compared to hybrid models, such as CMOD5.N and various PR models, CMODH achieves the smallest RMSE for wind speed retrieval. Moreover, for the first time, it is shown that the wind speed retrieval capability of CMODH is better than that of CMOD5.N for incidence angles between 30° and 49° and wind speeds between 10 and 20 m/s, which provides a new prospective on wind retrieval improvement using a combination of VV- and HH-polarized SAR observations.

A Geophysical Model Function for Wind Speed Retrieval From C-Band HH-Polarized Synthetic Aperture Radar

Synthetic aperture radar (SAR) imagery is routinely acquired at HH-polarization in high-latitude areas for measuring surface wind over the ocean. However, in the contrary of VV-polarization, there is no HH-polarization geophysical model function (GMF) exists to directly retrieve wind speed from SAR images. In general, HH-polarized normalized radar cross section (NRCS) is thus converted into VV-polarization and then conventional CMOD functions are used with auxiliary wind direction information for wind speed retrieval. In this letter, we propose a new GMF for SAR ocean surface wind speed retrieval, called CMODH, which relates the C-band NRCS acquired at HH-polarization over the ocean, to the 10-m height wind speed, incident angle, and relative wind direction. We first use more than 220 000 ENVISAT ASAR radar backscatter measurements collocated with ASCAT winds to derive the CMODH coefficients. Subsequently, 1459 RADARSAT-2 (RS-2) and 428 Sentinel-1A/B (Sl-1A/B) HH-polarized SAR acquisitions under different wind speeds are matched to in situ buoy observations to validate CMODH. The statistical comparisons between SAR-observed and simulated NRCS show a bias of −0.07 dB and a root-mean-square error of 1.62 dB for RS-2, and −0.01 dB and 2.48 dB for S1-1A/B. These results suggest that the proposed CMODH has the potential to directly retrieve ocean surface wind speeds using C-band SAR images acquired at HH-polarization, with no need for NRCS transformation by using various empirical and theoretical polarization ratio models.

Radar backscattering changes in Arctic sea ice from late summer to early autumn observed by space-borne X-band HH-polarization SAR

ABSTRACTMelt ponds are believed to play an important role in sea ice dynamics because they accelerate the melting of sea ice in the warmer spring and summer months. Additionally, they are known to absorb solar radiation rather than reflect it as the surrounding sea ice does. However, the size and distribution of melt ponds are highly variable, and thus, the contribution of melt ponds to sea ice melting should differ based on the maturity of the melt pond. Because of the harsh conditions of the Arctic, estimating the actual surface changes via in situ measurements and/or optical remote sensing data is difficult. In this study, we present a high-resolution time-series analysis of the short-term variation of sea ice and melt ponds over the Beaufort Sea using space-borne multispectral and synthetic aperture radar (SAR) images. A KOMPSAT-3 (Korea Multi-Purpose Satellite-3) optical image was used for an initial classification of the surface types, and 15 TerraSAR-X SAR images covering 46 days in the 2014 Arctic summer were used to perform a dense time-series analysis. The surface of the target sea ice was classified into six categories based on spectral characteristics. The temporal variation of the radar backscattering coefficient in each class exhibited a distinct pattern, which was closely related to surface changes. Overall, changes in the radar backscattering coefficient indicated dynamic surface changes, except over pressure ridges. All ice classes showed a two-step decrease in radar backscattering, whereas snow-covered ice surfaces exhibited far fewer changes compared to bare ice surfaces. The surfaces adjacent to ponds showed stronger negative decreases than other classes. The changes in dark melt pond classes presented a complex non-linear decrease, which differed from the stepwise decrease of blue melt ponds. These observations can be used for important modelling studies of surface melting/freezing rates and to infer the variation over large areas using remote sensing data.

Towards an Operational SAR-Based Rice Monitoring System in Asia: Examples from 13 Demonstration Sites across Asia in the RIICE Project

Rice is the most important food security crop in Asia. Information on its seasonal extent forms part of the national accounting of many Asian countries. Synthetic Aperture Radar (SAR) imagery is highly suitable for detecting lowland rice, especially in tropical and subtropical regions, where pervasive cloud cover in the rainy seasons precludes the use of optical imagery. Here, we present a simple, robust, rule-based classification for mapping rice area with regularly acquired, multi-temporal, X-band, HH-polarized SAR imagery and site-specific parameters for classification. The rules for rice detection are based on the well-studied temporal signature of rice from SAR backscatter and its relationship with crop stages. We also present a procedure for estimating the parameters based on “temporal feature descriptors” that concisely characterize the key information in the rice signatures in monitored field locations within each site. We demonstrate the robustness of the approach on a very large dataset. A total of 127 images across 13 footprints in six countries in Asia were obtained between October 2012, and April 2014, covering 4.78 m ha. More than 1900 in-season site visits were conducted across 228 monitoring locations in the footprints for classification purposes, and more than 1300 field observations were made for accuracy assessment. Some 1.6 m ha of rice were mapped with classification accuracies from 85% to 95% based on the parameters that were closely related to the observed temporal feature descriptors derived for each site. The 13 sites capture much of the diversity in water management, crop establishment and maturity in South and Southeast Asia. The study demonstrates the feasibility of rice detection at the national scale using multi-temporal SAR imagery with robust classification methods and parameters that are based on the knowledge of the temporal dynamics of the rice crop. We highlight the need for the development of an open-access library of temporal signatures, further investigation into temporal feature descriptors and better ancillary data to reduce the risk of misclassification with surfaces that have temporal backscatter dynamics similar to those of rice. We conclude with observations on the need to define appropriate SAR acquisition plans to support policies and decisions related to food security.

Baltic Sea Ice Concentration Estimation Based on C-Band Dual-Polarized SAR Data

High-resolution ice concentration information is required for navigation purposes, validating ice models, and data assimilation. The currently available operational ice concentration products are based on microwave radiometer data, and their typical resolution is several kilometers. We present an algorithm for the estimation of ice concentration based on dual-polarized (HH/HV) C-band synthetic aperture radar (SAR) data. The algorithm is based on the multilayer perceptron (MLP) neural network. Ice concentration estimated based on the HH channel is used as one MLP input, and the local incidence angle is used as another. The additional inputs are based on the HV channel. Digitized Finnish Ice Service ice charts, which were also used as the training data, the SAR-based HH-channel ice concentration, and the ice concentration based on a radiometer algorithm are used as reference data sets. The results for the dual-polarized algorithm show improvement compared to the algorithm based on HH-polarized SAR data only.

Application of RADARSAT imagery to grassland biophysical heterogeneity assessment

Heterogeneity, the degree of dissimilarity, is one of the most important and widely applicable concepts in ecology. It is highly related to ecosystem conditions and wildlife habitat. In this study, the feasibility of applying RADARSAT-1 HH polarization synthetic aperture radar (SAR) imagery to heterogeneity studies is tested at the Grasslands National Park (GNP) and surrounding pastures, Saskatchewan, Canada. A RADARSAT-1 HH standard mode image acquired on 7 August 2003 was complemented by fieldwork in June and July 2003. Height heterogeneity index, Shannon's index, and standard deviation of biophysical parameters were calculated based on field data. Parameters from texture analysis and standard deviation of the backscatter were correlated with biophysical parameters to measure grasslands biophysical heterogeneity. The results showed that different parameters had various abilities to detect field heterogeneity. Grey level co-occurrence matrix mean, correlation, dissimilarity, and contrast showed the ability to detect the variation of grass biomass and total biomass.

Correction of Surface Roughness and Topographic Effects on Airborne SAR in Mountainous Rangeland Areas

Synthetic aperture radar (SAR) has potential use as a means of measuring soil-water content from space or high altitude airborne platforms. However, the signal from other landscape features, such as topography, vegetation, and surface roughness, may overwhelm and confound the soil moisture signal. Previous researchers tried to correct for topographic effects using incident angle (θ) as the most significant landscape variable. Those corrections were not effective because of the poor correlation between the SAR backscatter coefficient (σ 0 ) and θ. This may be attributed to surface roughness, vegetation effects, or low resolution digital elevation data. In this article, we determined the effects of vegetation, surface roughness, and topography on σ 0 . By comparing the normalized difference vegetation index (NDVI) and the SAR-based vegetation index (SBVI), it was found that the sparse vegetation cover of sagebrush in the rangeland areas of the Reynolds Creek Experimental Watershed (RCEW) does not affect L band HH polarization SAR σ 0 . Root mean square (RMS) height of surface (h), determined using L band copolarized SAR σ 0 in an empirical algorithm (Dubois et al., 1995), helped in stratifying the study area into low and high surface roughness areas. The high surface roughness areas, determined using the algorithm, were found to match with very rocky and extremely rocky areas, classified using detailed soil type information in a Geographic Information System (GIS). The performance of a θ-based correction function improved significantly after incorporating h into the model.

Agricultural crop monitoring using airborne multi-spectral imagery and C-band synthetic aperture radar

Airborne optical multi-spectral and C-band HH-polarized Synthetic Aperture Radar (SAR) imagery were acquired in conjunction with contemporaneous ground-based measurements of various crop conditions (Leaf Area Index, canopy temperature, plant height) at a test site in southern Alberta, Canada on July 19-20, 1994. Data were acquired for a variety of crops (wheat, canola, peas and beans) and irrigation practices. A number of crop condition-imagery relationships were examined to determine whether the imagery could be used to measure the various crop condition parameters. A number of statistically significant correlations were found between the imagery and the crop condition parameters, and these correlations vary as a function of crop type, sensor and crop condition parameter. The results suggest that airborne remote sensing is well suited for measuring variations in crop conditions and that C-band SAR and multi-spectral imagery provide complementary information.

An evaluation of Landsat TM spectral data and SAR-derived textural information for lithological discrimination in the Red Sea Hills, Sudan

The effectiveness of spectral and textural information in the identification of surface rock types in an arid region, the Red Sea Hills of Sudan, is evaluated using spectral information from the six Landsat TM optical bands and textural features derived from Shuttle Imaging Radar-C (SIR-C) C-band HH polarization data. An initial classification is derived from Landsat TM data alone using three classification algorithms, Gaussian maximum likelihood, a multi-layer feed-forward neural network and a Kohonen self-organizing feature map (SOM), to generate lithological maps, with classification accuracy being measured using a confusion matrix approach. The feed-forward neural net produced the highest overall classification accuracy of 57 per cent and was, therefore, selected for the second experiment, in which texture measures from SIR-C C-band HH-polarized synthetic aperture radar (SAR) data are added to selected TM spectral features. Four methods of measuring texture are employed, based on the Fourier power spectrum, grey level co-occurrence matrix (GLCM), multi-fractal measures, and the multiplicative autoregressive random field (MAR) model. The use of textural information together with a subset of the TM spectral features leads to an increase in classification accuracy to almost 70 per cent. Both the MAR model and the GLCM matrix approach perform better than Fourier and multi-fractal based methods of texture characterization.

Characterizing forest stands with multi-incidence angle and multi-polarized SAR data

The Shuttle Imaging Radar-B (SIR-B) was flown in October 1984 on Space Shuttle Mission 41-G, and digital L-band HH-polarized synthetic aperture radar (SAR) data were obtained for designated test sites around the world. In the U.S., SIR-B data were obtained at incidence angles of 28°, 45°, and 58° for a forested test site in northern Florida. This SIR-B data set is unique in that it was the first time in the history of the U.S. space program that a set of digitally correlated multiple incidence angle radar data had been obtained from satellite altitude. Quantitative analysis of the SIR-B data has shown that several different cover type/density classes of forest cover as well as other land use classes can be discriminated as a function of radar return and incidence angle. Deciduous forest stands underlain by water have a particularly strong return in the 28° incidence angle data, but cannot be discriminated from the surrounding pine stands in the 58° data. Analysis of multi-polarized data obtained for the same test site from aircraft altitudes indicates that more effective discrimination between deciduous, pine, and other cover types can be achieved with like-polarized (HH or VV) data than can be obtained with the cross-polarized (HV or VH) data.

Analysis of synthetic aperture radar data acquired over a variety of land cover

The results of Synthetic Aperture Radar (SAR) measurements over Kershaw County, Sooth Carolina, using HH, HV, and VV polarization and two-incidence angle X-band airborne SAR system and over Baldwin County, Alabama, using HH polarization L-band Shuttle Imaging Radar (SIR-A) are presented. The X-band data indicate higher HH than VV radar return for cypress forest with standing water. Muttipolarization (HH, HV, and VV) data help delineate several land-cover types that are difficult to delineate by the single polarization (HH) data. The L-band data indicate that radar return signal strength is highly correlated with tree height or age for three types of pine forest. It is found that delineation of urban/residential from deciduous forest is significantly improved by the inclusion of Landsat multispectral scanner data.